Token A
Your pick
name hiddenCASHCAT
- 24h volume
- $3,157,360
- 24h transactions
- 7,658
- Price change 24h
- −7.96%
- Pooled liquidity
- $9.6M
- Volume ÷ liquidity
- 0.33×
deep poolTraded for months
WhitepaperHow the scoring works →
Is this token structurally a trap?
Apiro is a pre-trade filter API for trading bots and AI agents on Robinhood Chain. It answers that one question in the milliseconds before you commit capital — with a score, a verdict and itemized reasoning.
- Ten structural signals
- REST API & MCP server
- Exit risk, not price calls
Two real tokens · same day · Robinhood Chain
Your bot reads these three fields. Which one does it buy?
Token B
Your pick
name hiddenROBINTWINE
- 24h volume
- $662,519
- 24h transactions
- 4,821
- Price change 24h
- +7.30%
- Pooled liquidity
- $5,622
- Volume ÷ liquidity
- 117.8×
criticalscore 0 · tradeable: false
- 117.8×daily volume against Token B’s entire pool
- 235×volume against its $2,817 market cap
- 0trades in an hour, seventeen hours later
The problem
Raw market data describes traps and healthy tokens identically.
Price, volume, liquidity, transaction counts — the fields every bot reads are accurate. On their own they’re close to useless for telling a viable trade from a designed loss, and they often make the dangerous token look more attractive, because manufactured volume reads as demand.
Turnover · Token B
$662,519÷$5,622=117.8×
Daily volume against pooled liquidity. A healthy token recycles its pool depth somewhere under three times a day. No pool can honestly recycle its entire depth 117 times.
Volume vs. market cap · Token B
$662,519÷$2,817=235×
Every coin in existence would have had to change hands 235 times in a single day to produce that volume honestly.
Seventeen hours later
4,819 trades in 24h→0in the last hour
The activity switched off. The price never crashed — it’s still nominally up. There is simply no longer a bid, so anyone holding it owns a position they can’t sell at any size.
This isn’t an exotic failure mode. It’s the dominant one — and the ratios that expose it are trivial arithmetic almost nobody performs at decision time.
Why more data isn’t the answer
Structural signals don’t decay when others use them.
Raw trade and liquidity data for Robinhood Chain is already a commodity — every major indexer covers it. A better feed gets you the same fields. A signal feed is worse.
Directional signals decay with adoption
Tell a thousand subscribers a whale just bought, and they race each other into the same position. The edge is consumed by the race — and it adds slippage to the very trades it recommends.
Structural facts stay true for everyone
A $5,622 pool behind $662,519 of volume is as true for the thousandth subscriber as the first. Nobody’s avoidance of a trap makes it less of a trap — so there’s no race.
Value per subscriber as adoption grows
IllustrativeCommodity layer
Market data indexers
Price, volume, liquidity and transaction counts. Accurate, widely available — and insufficient on their own.
Judgement layer
Apiro
Consumes commodity data and sells judgement: ten structural signals cross-referenced into a score, a verdict and plain-language reasons.
Execution layer
Your bot or agent
Skips, sizes down or proceeds — before an order is ever submitted.
Methodology
Ten structural signals. No privileged access.
Every signal is computed from public DEX pool state — no latency advantage, no co-located node. Each is scored on a severity band and adds a weighted penalty to a score that starts at 100 and descends.
01thin_liquidity
Liquidity depth
Pooled USD across every pool. Below roughly $50k, a meaningful exit moves the price against you. Below $10k, exit is effectively theoretical.
Token B: $5,622 pooled
02extreme_turnover
Turnover ratio
24h volume ÷ pooled liquidity. Healthy tokens recycle their depth under 3× a day; above 20× isn’t achievable by organic trading. The highest-signal metric in the system.
03volume_exceeds_mcap
Volume vs. market cap
24h volume ÷ market cap. Above 10×, every token in existence would need to trade ten times over in a day.
Failed: 160.2× · 235.2×
04newly_launched
Pool age
Hours since the earliest pool was created. Not a claim that new tokens are scams — an irreducible unknown while the deployer keeps every option.
Token B: 16.8 h old
05price_collapse
Price collapse
A 24h drawdown beyond 50% on a token only days old is usually not a dip — it’s the event itself, in progress.
Watches for >50% drawdowns
06sell_pressure
Sell pressure
Sells as a share of all transactions. Sustained imbalance beyond 70% indicates distribution, not accumulation.
Watches for >70% sells
07micro_trade_signature
Micro-trade signature
Thousands of near-uniform trades averaging under $20 describe a machine, not a market. Human demand doesn’t distribute that way.
Watches for <$20 average trades
08price_dispersion
Cross-pool dispersion
Arbitrage closes price gaps between pools within blocks. A persistent spread means your own trade will be priced just as badly.
Persistent spread = too thin to arb
09activity_cessation
Activity cessation
Last-hour trades against the last 24 hours. Added after live testing exposed a state the other nine couldn’t see: six figures of daily volume, and not a single trade in the last hour.
4,819trades · prior 24h
0trades · last hour
10liquidity_concentration
Liquidity concentration
Share of liquidity held in a single pool. One withdrawal can remove the entire market.
One pool = one point of failure
Live example
One call. A score, a verdict — and the reasons.
Every flag carries the value that triggered it and a plain-language explanation. Rule-based bots threshold on the number; reasoning agents weigh the prose against a user’s risk tolerance.
GET/v1/tokens/{address}/risk
This demo replays a recorded response for ROBINTWINE from live Robinhood Chain data, abbreviated as published in the Apiro whitepaper. Endpoint domains are indicative ahead of launch.
Integration
Drops in right before order submission.
Signals move slowly, so responses are safely cacheable for a minute or more — negligible latency on a hot path, negligible cost for high-frequency strategies.
One call, one branch.
Place the tradeable gate immediately before order submission. It returns a single boolean plus the flags that blocked the trade — no report parsing required.
- Choose your own
min_scorethreshold - Cache responses for a minute or more
- Compare outcomes against your unfiltered baseline
import httpx
def should_trade(address: str, min_score: int = 45) -> bool:
r = httpx.get(
f"https://api.apiro.io/v1/tokens/{address}/tradeable",
params={"min_score": min_score},
headers={"Authorization": f"Bearer {API_KEY}"},
timeout=2.0,
)
return r.json()["tradeable"]
if should_trade(token_address):
execute_buy(token_address, size)Let your agent check what it’s about to trade.
Apiro ships as an MCP server, so it drops into the same configuration as your agent’s trading tools — including Robinhood’s Agentic Trading, which connects third-party agents over MCP. The agent holds both the ability to trade and the ability to check.
assess_token_risk — structural assessment of one token before trading it.screen_tokens — up to 30 candidates, ranked safest-first with blocking flags.{
"mcpServers": {
"apiro": {
"url": "https://mcp.apiro.io/sse",
"headers": { "Authorization": "Bearer YOUR_KEY" }
}
}
}Filter candidate sets before deeper analysis runs.
New-launch monitors and trending scans generate far more candidates than they can evaluate. Batch scoring removes the majority up front, at a fraction of the cost of checking each one individually.
- Up to 30 addresses per call
- Sort on
raw_penaltyto rank below the critical threshold - Agents get the same via
screen_tokens
curl -X POST "https://api.apiro.io/v1/tokens/risk" \
-H "Authorization: Bearer $APIRO_API_KEY" \
-H "Content-Type: application/json" \
-d '{"addresses": ["0xC7110dd6343fb126865DB673284D5B67F247470F"]}'Design principles
Built to be trusted by machines — and by the people behind them.
Structural, not directional
Apiro never tells you what to buy. A bad score is a reason to skip or size down — explicitly not a short signal. The claim is about exit risk, not price path.
Slow signals, deliberately
Being a few seconds behind costs nothing on structural facts. Apiro recomputes on a cadence of minutes, stays out of the latency arms race, and passes the savings through as pricing.
Explanations, not just numbers
Every flag carries its triggering value and why it matters. Bots ignore the prose and threshold on the number; agents use the prose to weigh a finding against your instructions.
Honest about proxies
Manufactured-volume signals are statistical patterns, not proof of intent. Apiro says “consistent with” rather than asserting fraud — and returns its own caveats in every response.
Pricing
Flat by tier. Never metered per call.
Per-call billing is hostile to strategies on thin margins — and discourages exactly the defensive checking Apiro exists to encourage.
Free
Evaluation and low-volume bots
$0
- Rate-limited risk scoring
- Full flag detail
- Community support
Developer
Individual bot operators
Flat rateAnnounced at launch
- Higher limits
- Batch endpoint
- MCP access
- Webhooks Planned
Professional
Trading desks, multi-agent operators
Flat rateAnnounced at launch
- Production limits
- SLA & priority support
- Historical scoring Planned
- Deployer graph Planned
Enterprise
Platform integrators
Custom
- Custom limits
- Dedicated infrastructure
- White-label
- Direct support channel
Indicative structure, subject to change at launch. Compare tiers in detail
FAQ
What bot builders ask first.
Does Apiro predict which tokens will go up?
No — and any product claiming to do so reliably is either mistaken or selling something. Apiro identifies exit risk. A token can score well and still fall; the score describes whether you can get out, not where the price goes.
Why not just check liquidity myself?
You can, and you should. Apiro’s value is the combination — ten signals computed consistently, cross-referenced, updated continuously, and delivered in a form your bot or agent consumes in one call.
The turnover and volume-to-market-cap ratios in particular are trivial arithmetic that almost nobody performs at decision time, which is exactly why manufactured volume works as well as it does.
How is this different from a token scanner?
General EVM scanners focus on contract-level traps — honeypots, mint functions — across many chains. Apiro focuses on market-structure traps specific to Robinhood Chain’s dynamics. The two are complementary, and contract-level analysis is on the roadmap precisely because the combination is stronger than either alone.
What if the data source is wrong or stale?
The activity_cessation signal exists partly to catch this. When a feed’s figures stop moving while still claiming high volume, that state is surfaced rather than silently passed through as a healthy market.
Can I run it against other chains?
The methodology generalises to any DEX-based chain. Robinhood Chain is the initial focus because it’s new enough that the tooling gap is real and specific.
Get started
From API key to a filtered strategy in four steps.
Request an API key
Start on the free tier — full flag detail is included.
Score a token you already have an opinion about
It’s the fastest way to calibrate how much to trust the output.
Gate one strategy
Add the
tradeablecheck to a single strategy and compare outcomes against your unfiltered baseline.Wire up your agents
For agent-based trading, add the MCP endpoint alongside your existing trading tools.
Request an API key
Tell us what you’re building and we’ll follow up with access details.